I need to get the standard errors of coefficients with Ridge Regression, by calculating the SE of the beta estimates after I choose the right lambda.

ridfit = lm.ridge(y ~ atemp+aconc+wtemp+sconc+ableach, data = white, lambda = seq(0,10,.1))
## modified HKB estimator is 4.188714 
designmat = cbind(rep(1,length=length(y)),atemp, aconc,wtemp,sconc,ableach)
ridfitHKB = lm.ridge(y ~ atemp+aconc+wtemp+sconc+ableach, data = white, lambda = 4.188714)
fvr = designmat%*%c(-18.0093065, 0.6419894, 17.4223497, 0.6830183, 22.0336205, 14.8206600)
sse.rid = sum((y-fvr)^2)

Essentially, I can find the SSE of the ridge regression model, but I need to find the SE for each Bi. I am to use the lambda value selected by the above select(ridfit), which = 4.188714. But I do not know how to calculate the Bi SE with only that information.

I know that the code will not help anyone in computation, but I suspect showing my methodology and noting that I am to compute the Standard Errors from using the lambda value chosen will help.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.